* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Vacuum Ultraviolet Spectroscopy and Photochemistry of Zinc
State of matter wikipedia , lookup
Chemical bond wikipedia , lookup
Rutherford backscattering spectrometry wikipedia , lookup
Auger electron spectroscopy wikipedia , lookup
Two-dimensional nuclear magnetic resonance spectroscopy wikipedia , lookup
Atomic orbital wikipedia , lookup
Chemical imaging wikipedia , lookup
Physical organic chemistry wikipedia , lookup
Heat transfer physics wikipedia , lookup
Rotational spectroscopy wikipedia , lookup
Molecular orbital wikipedia , lookup
X-ray fluorescence wikipedia , lookup
Electron configuration wikipedia , lookup
Metastable inner-shell molecular state wikipedia , lookup
Rotational–vibrational spectroscopy wikipedia , lookup
Mössbauer spectroscopy wikipedia , lookup
Ultraviolet–visible spectroscopy wikipedia , lookup
Astronomical spectroscopy wikipedia , lookup
Franck–Condon principle wikipedia , lookup
Article pubs.acs.org/JPCA Vacuum Ultraviolet Spectroscopy and Photochemistry of Zinc Dihydride and Related Molecules in Low-Temperature Matrices Chris Henchy, Una Kilmartin, and John G. McCaffrey* Department of Chemistry, National University of Ireland Maynooth, Maynooth, Co. Kildare, Ireland S Supporting Information * ABSTRACT: Optical absorption spectra of thin film samples, formed by the codeposition of zinc vapor with D2 and CH4, have been recorded with synchrotron radiation. With sufficiently low metal vapor flux, samples deposited at 4 K were found to consist exclusively of isolated zinc atoms for both solids. The atomic absorption bands in the quantum solids D2 and CH4 were found to exhibit large bandwidths, behavior related to the high lattice frequencies of these low mass solids. The reactivity of atomic zinc was promoted with 1 P state photolysis leading to the first recording of electronic absorption spectra for the molecules ZnD2 and CH3ZnH in the vacuum ultraviolet (VUV) region. 3P state luminescence of atomic zinc observed in the Zn/CH4 system points to the involvement of the spin triplet state in the relaxation of CH3ZnH system as it evolves into the C3v ground state. This state is not involved in the relaxation of the higher symmetry molecule ZnD2. Time dependent density functional theory (TD-DFT) calculations were conducted to predict the electronic transitions of the inserted molecular species. Comparisons with experimental data indicate the predicted transition energies are approximately 0.5 eV less than the recorded values. Possible reasons for the discrepancy are discussed. The molecular photochemistry of ZnD2 and CH3ZnH observed in the VUV was modeled successfully with a simple four-valence electron AH2 Walsh-type diagram. I. INTRODUCTION Zinc dihydride (ZnH2) has been the subject of a number of investigations in recent years. Two primary reasons can be identified for the current interest in this triatomic molecule. The first is that ZnH2 can be considered the parent molecule for the Group 12 metal M(ns)2 dihydrides and considerable interest thereby exists for the determination of its fundamental molecular constants. The second stems from the rich photochemistry of the related dialkyl metals, MR2, such as dimethylzinc. Of particular significance in obtaining precise molecular constants for ZnH2 was the high-resolution mid-IR emission study conducted by Bernath and co-workers.1 Analysis of rotationally resolved gas phase spectra for ZnH2 and ZnD2 allowed extraction of the equilibrium bond lengths as 1.52413 and 1.52394 Å, respectively, of these two 64Zn-containing isotopologues. The band origins of the asymmetric stretching vibrations for 64ZnH2 and 64ZnD2 were identified at 1889.433 and 1371.631 cm−1, respectively. In addition, verification of the expected linear, centro-symmetric geometry was also achieved. Quite recently, Sebald et al.2 have conducted high level ab initio calculations to generate an accurate potential energy function for the ground X 1Σg+ electronic state of ZnH2. The coupledcluster method utilized in the theoretical study yielded molecular constants in excellent agreement with Bernath’s experimental values for the three hydrogenic isotopologues up to approximately 15 000 cm−1 above the equilibrium minimum. Recently microwave absorption spectroscopy3 has been used to determine the structure and bonding in methylzinc hydride © 2013 American Chemical Society (CH3ZnH). The Zn−H bond length was determined to be 1.521 Å, very similar to that of ZnH2. Consistent with the linear structure of zinc dihydride, the geometry of CH3ZnH was found to be C3v. The vibrational spectroscopy of zinc dihydride has been recorded in a variety of matrix-isolation experiments. Wang and Andrews4 generated this species with broadband photolysis of atomic zinc in solid para-H2 (and H2-doped neon) observing the asymmetric stretching mode at 1880.5 (1880.6) cm−1 and the bending mode at 631.9 (632.5) cm−1 for 64ZnH2. The two other hydrogen isotopologues were also reported in that study, and results were compared with DFT predictions. These modes were reported previously by Andrews5 as 1870.8 and 628.3 cm−1 for ZnH2 in argon matrices. Our group6 also observed them at this position in Ar but generated with secondary photolysis of dimethylzinc. Quantum chemical calculations have been used at a variety of levels of sophistication to predict the bond length and vibrational frequencies of ground state zinc dihydride. A comparison with Bernath’s experimental values in Table S1 reveals that DFT calculations conducted with the B3LYP functional and the 6-311++G(3df,3pd) basis set provide results in good agreement with observations. This basis set has been used extensively in the present work, yielding results in Received: May 13, 2013 Revised: August 23, 2013 Published: August 23, 2013 9168 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178 The Journal of Physical Chemistry A Article complete agreement with the earlier1 DFT findings. Thus the ground state equilibrium bond length was found to be 1.5413 Å, and the fundamental vibrational frequencies of ν1, the symmetric (σg), and ν3, the asymmetric (σu) stretches were 1915.5 and 1926.8 cm−1, respectively, while ν2, the bending (πu) mode, is predicted at 642.8 cm−1 for 64ZnH2. Table S1 in the Supporting Information collects these DFT harmonic data and 64ZnD2 results for comparison with a variety of experimental IR findings both in the gas phase and matrices. Detailed gas phase collisional deactivation studies have been conducted by Breckenridge and co-workers7 on the excited triplet 4p 3P1 state of atomic zinc8 with pump−probe methods, for molecular hydrogen and its isotopologues, HD and D2. More recently Umemoto and co-workers9 have extended this work to the excited singlet 4p 1P1 state. Analysis of the rotational distribution in the zinc hydride product molecule has allowed the inference of a long-lived intermediate in these reactions. This proposal has been examined by Novaro’s group10 who have determined the reaction paths of excited state atomic zinc with molecular hydrogen. The findings of this theoretical work largely confirm the existence of a deeply bound linear inserted H−Zn−H species, but involving an endothermic reaction with respect to the separated species ground (1S0) state atomic zinc and molecular hydrogen. In the present contribution we examine the electronic spectroscopy of the zinc dihydrides (ZnD2 and ZnH2) as well as methylzinc hydride (CH3ZnH) generated from the reaction of excited 1P1 state atomic zinc with D2/H2 and CH4. This is achieved for atomic zinc isolated in solid matrices formed from the co-condensation of zinc vapor with molecular deuterium (D2), methane (CH4), and H2 in doped Ar matrices. Any luminescence detected in the current reactive host solids as a result of zinc 1P1 photoexcitation is compared with that previously observed by our group for atomic zinc isolated in inert solids such as neon11 and argon12 matrices. The electronic spectra of the resulting “inserted” species have been recorded for the first time in the vacuum-UV range. Furthermore, repeated scans in the VUV region have revealed photodissociation of the inserted molecular species regenerating atomic zinc. Time-dependent density function theory (TDDFT) calculations were then used to analyze the nature of the electronic transitions responsible for the absorption spectra of the inserted molecules. This method was also used to examine the efficient photodissociation observed for these species in the VUV. a 0.25 m Spex 240 M imaging monochromator which was coupled by a 2 m optical fiber to the sample. The CCD chip consists of 1100 × 330 pixels and was operated at −100 °C with exposure times typically of 5 min. Samples were made by the codeposition of zinc vapor with a large excess of the host gas (deuterium or methane) onto an LiF window at 4 K. This was the minimum temperature attainable on the sample holder, with the Leybold−Heraeus flow-through liquid helium cryostat used. All of the host materials used in the present study, Ne, Ar, D2, H2, and CH4, are transparent in the wavelength range of interest with the exception of CH4, which starts to absorb for λ < 145 nm. Zinc vapor was produced by electron bombardment of a 2.0 mm Zn rod placed in a molybdenum holder. Full details of the electron bombardment metal atom source used have been presented elsewhere.13 In contrast to solid D2, none of the attempts made at 4 K to isolate zinc atoms in neat, solid hydrogen were successful. However, Ar samples containing up to 10% H2 were found to be stable even when deposited at 12 K. Zn/H2 photochemistry was analyzed in these H2-doped Ar matrices. B. Calculations. The DFT calculations conducted in this study all used the B3LYP functional and were performed with the Gaussian-03 suite of programs14 running on a Linux workstation, which has been described15 in our earlier work. In the present study the 6-311++G (3df,3pd) basis set was used exclusively. As indicated in Table 1, this method yielded an optimized geometry and vibrational frequencies of ZnH2 identical to the values previously published in Table S13 of the supplementary data of ref 1. Moreover, the DFT predictions obtained for CH3ZnH also agreed with earlier DFT,16 microwave,3 and matrix results5 as listed in Table S1 of the Supporting Information (SI). Table 1. A Comparison of the Structural and Vibrational Data of CH3−Zn-H Predicted by the Current DFT Calculations and Determined in Previous Experimental Studiesa DFT/B3LYP Zn−H Zn−C C−H Zn−C−H H−C−H H−Zn−C II. METHODS A. Experimental Section. Absorption spectra in the vacuum ultraviolet (VUV) region were recorded using synchrotron radiation with the HIGITI setup on the W3.1 beamline at HASYLAB, DESY, Hamburg, Germany. Scans in the 120−350 nm range were done with an LiF filter in the primary monochromatora modified 1.0 m Wadsworthto preclude short wavelength light reaching the sample via high order transmittance. A sodium-salicylate converter was used with an XP2020 photomultiplier tube (PMT) for light detection in the VUV. Emission spectra in this region were recorded with a 0.4 m Seya-Namioka vacuum monochromator using a Hamamatsu (model 1645U-09) multichannel plate for photon detection. A liquid nitrogen-cooled, charged coupled device (CCD) detector (Princeton Instruments, model LN/ CCD 1108PB) was used to record emission spectra in the UV− visible region. This steady-state CCD camera was mounted on C−H s str H−Zn str CH3 s bend C−Zn str C−H as.str CH3 as bend CH3 rock C−Zn−H bend experimental results Bond Length (Å) 1.5425 1.9546 1.0912 Bond Angle (deg) 110.9 107.9 180.0 Frequencies (cm−1), A1 modes 3021.1 1902.7 1208.4 552.3 Frequencies (cm−1), E modes 3094.2 1456.4 702.9 422.9 1.521 1.928 1.14 110.2 180.0 2920 1866 1179 564 687 443 a The structural experimental data are from gas phase microwave spectra presented in ref 3, while the vibrational data are from matrix results presented in ref 5. The DFT vibrational frequencies quoted are harmonic values. All of the vibrational frequencies listed are in wavenumber (cm−1) units, while the bond lengths are in Angstrom units. 9169 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178 The Journal of Physical Chemistry A Article dashed vertical line in Figure 1, a blue shift of the band occurs from the 213.9 nm gas phase position17 of the atomic transition. Two other major characteristics of the atomic absorption profile recorded in solid D2 are the large bandwidth (∼2050 cm−1) and the pronounced 3-fold splitting with resolved peaks at 206.4, 209.6, and 213.1 nm. Shown by the middle trace in Figure 1 is the absorption spectrum recorded for zinc isolated in a pure methane matrix (Zn/CH4). In contrast to the deuterium sample, the absorption band is now centered slightly to the red of the gas phase transition and is considerably narrower. However, it also exhibits characteristic 3-fold splitting, with partially resolved features at 212.6, 214.5, and 216.5 nm. For comparison purposes, the absorption spectrum of atomic zinc isolated in solid neon11 is shown by the bottom trace in Figure 1. While the atomic band in neon is located blue of the gas phase transitionroughly in the same region as Zn/D23-fold splitting is not evident and additional bands due to Zn dimer are now present at the indicated positions. The existence of pronounced dimer bands in neon samples reveals more efficient isolation of zinc atoms in solid deuterium and methane matrices than in neon. This difference will be discussed in more detail ahead in relation to possible site occupancies. Lineshape analyses done on the absorption bands of the Zn/ D2, Zn/CH4, and Zn/Ar systems are presented together in Figure S1 of the Supporting Information. In the Gaussian fits shown in Figure S1 three components, all with the same line width, were used on a given system. With the exception of Zn/ CH4, where a minor blue site exits, the use of only three components provides a good description of the recorded bands. The same fitting procedure was successful for all of the systems, the details of which are collected in Table 2. This indicates that similar 3-fold splitting of the 4p 1P1 ← 4s 1S0 electronic transition occurs in the three hosts. As the Zn/Ne absorption and excitation spectra have recently been analyzed11 in detail, only the relevant aspects will be compared with the present data. The fits done on Zn/Ne utilized different parameters,11 including variable linewidths and the requirement of a fourth Excited state calculations on ZnH2 (CH3ZnH) were carried out with the TD-DFT method using the optimized geometry of the D∞h (C3v) ground state molecule. Once calculations had converged, the excitation energies, oscillator strengths (f), and transition dipole moments of the first 50 excited states were extracted. The energies of the molecular orbitals were obtained using the “pop=reg” command, yielding values for all 16 occupied and 101 virtual orbitals of ZnH2. Excited state calculations on methylzinc hydride (CH3ZnH) were conducted in a similar fashion but in this case involving 20 occupied and 172 virtual orbitals. The molecular orbital energies were also determined as a function of angle between the linear and bent (90°) geometries. By doing these calculations in 5° increments, so-called “Walsh” diagrams were thereby produced. The MO plots shown in these diagrams were produced from the electronic density. This was done in Gaussian 03 by generating a “cube” file for a given molecular orbital. The electronic density was then plotted as a wireframe over the molecule showing the shape and parity of the orbitals of interest. III. RESULTS A. Deposition. An absorption spectrum recorded in the VUV-UV region for a Zn/D2 sample formed with low metal loading is shown by the top trace in Figure 1. Only a single, albeit structured, band is present whose center feature is located at 209.6 nm. Due to the sparsity of electronic transitions of atomic zinc in the wavelength range longer than 200 nm, this 3fold split absorption band can immediately be assigned to the resonance 4p 1P1 ← 4s 1S0 transition. As indicated by the Table 2. Absorption Band Positions Recorded for the 4p 1 P1−4s 1S0 Transition of Atomic Zinc Isolated in the Solids Formed from the Light Materials Deuterium, Methane, Neon, and Argon at 4.2 Ka Zn D2 CH4 Ne Ar Figure 1. Absorption spectra recorded for Zn isolated in the molecular matrices D2 and CH4 deposited at 4 and 12 K, respectively, both of which were scanned at 4.2 K. The location of the resonance 4p 1P1 ← 4s 1S0 transition of atomic zinc in the gas phase (213.9 nm) is indicated by the dashed vertical line. As indicated in the figure, a blue shift of the entire band occurs for Zn/D2 from the gas phase position, but a red shift exists in Zn/CH4. For comparison the absorption spectrum of Zn/Ne is presented on the bottom revealing a less resolved main band and features recently attributed (ref 11) to zinc dimer. absorption center (nm) absorption energy (cm−1) Δ (cm−1) 206.4 209.6 213.1 212.6 214.5 216.5 203.68 205.17 206.15 205.15 206.75 208.47 48444 47700 46992 47025 46614 46191 49095 48740 48509 48747 48367 47967 742 742 742 435 435 435 600 561 375 412 412 412 The gas phase position17 of this transition is at 46745.413 cm−1. All of the matrix energies quoted are in wavenumber (cm−1) units. The bandwidths (Δ, fwhm) quoted for the matrix features were extracted in the Gaussian lineshape fits of the spectra presented in Figure S1. The quoted values were calculated as Δ = s√[8 ln(2)] where s is the variance in the fitted function, G(x) = a exp[−(x − m)2/(2s2)]. a 9170 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178 The Journal of Physical Chemistry A Article component. This contrasting behavior points to different site occupancy in solid neon compared with the molecular hosts studied in the present work and the earlier12 Ar data. B. Luminescence. No emission was observed for Zn/D2 samples either in the UV or in the visible spectral regions with excitation into the atomic resonance absorption band shown in Figure 1. This situation is depicted in the central panel of Figure 2 with excitation at 210 nm. It is in stark contrast to Figure 2. A summary of the emission spectra recorded at 4.2 K in the Zn/D2 and Zn/CH4 systems. As indicated by the middle trace, no emission exists in the Zn/D2 system, while only the spin triplet is present (bottom trace) for Zn/CH4. It is located at the red-shift of the atomic zinc 4p1 3P1−4s 1S0 transition which occurs in the gas phase at approximately 308 nm. For comparison the emission recorded in Zn/ Ne is shown in which singlet state atomic emission is observed. Figure 3. Changes observed in the VUV−UV absorption spectra of the Zn/D2 system upon atomic photolysis at specific wavelengths (indicated by the arrows in Figure 1) shown as difference spectra. The negative peaks in the plots indicate the loss of the atomic bands, while the positive ones reveal the growth of transitions in the VUV. In addition to the pure D2 matrix result, the response of a 10% D2 in Ar sample and a 10% H2 in Ar sample to similar irradiation are shown in the lower panel. The dip evident at approximately 185 nm in the 10% D2 in Ar sample arose due to a change in the SR beam position during the scan and is not of photochemical significance. atomic zinc isolated in the solid methane where, as shown in the bottom panel, singlet excitation at 214 nm produces emission at 348 nm and a red shoulder at 380 nm. From the long-lived nature18 of this near-UV emission, these bands can be attributed to the atomic Zn 4p 3P1 state which occurs in the gas phase17 at 307.68 nm. Comparable excitation at 205.4 nm in Zn/Ne produces11 intense emission at 212.8 nm as shown in the top trace of Figure 2. With a radiative lifetime of 1.16 ns and its spectral location, these characteristics indicate that this emission can be assigned to the fully allowed 4p 1P1 → 4s 1S0 transition. Thus, singlet P-state fluorescence dominates in neon,11 as it does in argon and krypton matrices.12 In light of the data presented in Figure 2, it is thereby evident that singlet fluorescence of atomic zinc is completely quenched in both solid D2 and CH4. In the latter system, the atomic triplet state is however produced with high efficiency, a situation observed previously in Zn/Xe. C. Atomic Photolysis. The reactivity of excited 4p 1P1 state atomic zinc in solid deuterium was monitored in the VUV upon photolysis at 210 nm, a wavelength indicated by the upper arrow in Figure 1. The result of a 15 min irradiation is presented as a “difference” absorption spectrum in the top trace of Figure 3. In this spectrum depletion of the 3-fold-split atomic resonance band centered at 210 nm is very evident, but more significantly, this depletion is accompanied by the appearance of a pair of broad, featureless bands at 145 and 155 nm as well as the hint of a much weaker band at 168 nm. Because of the absence of samples composed of atomic zinc isolated in neat H2 matrices, the photochemical activity of zinc with hydrogen was examined in Ar samples containing 10% H2 (Zn/H2Ar). To allow direct comparison with the results obtained for the pure Zn/D2 system, D2-doped Ar samples of composition similar to H2Ar were also produced. The effect of atomic zinc photolysis in a Zn/D2Ar sample is shown in the middle trace of Figure 3. As observed with pure D2 samples (upper trace), new absorption bands are produced in the 140− 160 nm region, but the changes produced in the atomic absorption in the doped D2/Ar samples are more complex. In particular, an increase is observed in one feature located at 215 nm which is due to photolytically induced site interconversion, that is, the isolation of the zinc atom in a new site in the D2doped Ar sample. A similar, but more pronounced, effect is evident in the difference spectrum recorded for a Zn/10%H2Ar sample, shown in the bottom trace in Figure 3. This effect arises from ill-defined sites of isolation of atomic zinc in the lessstructured doped-Ar samples. Clearly the VUV absorptions of Zn/10%H2Ar occur in the same spectral region but are less structured than in the corresponding D2-doped Ar samples. It is also evident that the amount of photochemical product generated in the H2-doped Ar samples is less than in the equivalent D2-doped Ar samples. This is probably related to the less efficient trapping of the lighter isotopologue in Ar. Excited state reactivity of zinc with methane was examined with photolysis at 214 nm whose location is shown by the 9171 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178 The Journal of Physical Chemistry A Article photolysis. As indicated in the two lower traces in Figure 5, irradiation at 145 and 153 nm into each excited molecular band arrow in the middle panel of Figure 1. The result of a 15 min irradiation at this wavelength is presented as a “difference” absorption spectrum in the upper trace of Figure 4. In this Figure 5. VUV photodissociation of ZnD2 isolated in solid D2 revealing, in the lower two traces, the regeneration of atomic zinc and the depletion of the molecular bands. Figure 4. Same as for Figure 3 except for the Zn/CH4 system. The lower trace shows the results obtained for a 5% methane in the Ar sample. resulted in the depletion of both. This depletion was accompanied by the regeneration of the characteristic atomic zinc absorption at 210 nm. After regeneration of atomic zinc, the production of the inserted molecule was shown to be possible once again with atomic photolysis. Thus the insertion/ dissociation cycle could be repeated several times. As a result of this and the characteristic atomic zinc absorption band shape, it is highly likely the other product of the photodissociation is molecular hydrogen and not hydrogen atoms. At the very low temperatures this work was done, some isolation of H atoms should be possible in the solid hosts. The absence of H atoms is indicative of a concerted dissociative mechanism whereby both Zn−H bonds break at the same time, leading to the formation of an H−H bond. In the Zn/CH4 system all three VUV molecular bands were found to be photochemically active. Irradiation into any of them resulted (data not shown) in the regeneration of the atomic zinc absorption at 214 nm and depletion of the new molecular bands. This behavior is similar to that observed by us previously19 in the photochemistry of matrix-isolated dimethylzinc (DMZ). However, in contrast to the dissociation of DMZ, where secondary photolysis never regenerated the starting material, methylzinc hydride species could be regenerated with atomic photolysis. This difference stems from the inability of metal atoms to insert into the C−C bond of ethane compared to their facile insertion into the C−H bonds. E. TD-DFT Calculations. The electronic absorptions of ZnH2 predicted by TD-DFT calculations are collected in Table 3 for the singlet−singlet transitions. In this table it is evident that the lowest energy transition is from the HOMO (MO 16) spectrum depletion of the atomic resonance band is evident as well as the appearance of at least three featureless bands at 174, 162, and 149 nm. These bands are all in the vacuum-UV but at lower energy than those observed in Zn/D2. The noise evident at wavelengths shorter than 145 nm in the Zn/CH4 spectrum is due to the strong absorption of the host material, methane, in this region. The reactivity of excited state atomic zinc with methane was also examined in CH4 doped-Ar matrices. The “difference” absorption spectrum shown by the bottom trace in Figure 4 presents the results of a 15 min irradiation at 208 nm. In this 5% CH4/Ar sample, depletion of the atomic resonance band is much simpler than what is observed in the H2-doped Ar samples with clean loss of the atomic features, that is, free from site interconversion. The appearance of at least three featureless bands in the VUV is also evident in the figure with a better S/N ratio than obtained in the pure methane sample. However, as shown by the gray trace in the upper panel of Figure S2, the three new bands in Ar are all to the blue of the three bands in pure methane (black trace). The locations of the bands in doped-CH4 Ar samples are 168, 156, and 146.5 nm compared with 174, 162, and 149 nm in neat CH4. In contrast, the bottom panel of Figure S2 reveals that little or no shift occurs in the bands observed in the Zn/D2 system in pure deuterium (red trace) and D2-doped Ar matrices (gray trace). The contrasting behavior of the VUV bands of ZnD2 and CH3ZnH in Ar may be related to site occupancy whereby the latter is cramped in a divacancy, while the former fits easily into such a site. D. VUV Photolysis. Both of the strong vacuum-UV bands in the Zn/D2 system were found to be very sensitive to 9172 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178 The Journal of Physical Chemistry A Article Table 3. Excitation Energies and Oscillator Strengths for Spin-Allowed Transitions from the Ground X 1Σg State of ZnH2a state no. 1 2 3 4 5 6 7 8 9 10 11 orbitals/coeff. 16 16 16 16 16 16 15 15 15 15 15 16 16 16 15 16 15 15 15 15 15 15 16 16 → → → → → → → → → → → → → → → → → → → → → → → → 17/0.31058 18/0.62335 17/0.62335 18/−0.31058 19/0.66612 20/−0.23178 17/0.10422 18/0.66126 17/0.66126 18/−0.10422 21/−0.10725 19/0.21522 20/0.58930 26/−0.18648 20/−0.10301 21/0.68731 19/0.25477 20/0.66992 21/0.10189 21/−0.40793 22/0.56005 21/0.56142 20/0.13779 22/0.35017 state sym. E (eV) λ (nm) E (cm−1) f Πg 5.7997 213.78 46777 0.0000 Πg 5.7997 213.78 46777 0.0000 Σu 7.2816 170.27 58730 0.0910 Πu 8.0394 154.22 64842 0.3502 Πu 8.0394 154.22 64842 0.3502 Σu 8.1768 151.63 65950 0.7040 Σg 8.3336 148.78 67213 0.0000 Σg Σg 9.0256 9.3425 137.37 132.71 72796 75352 0.0000 0.0000 Σu 9.8924 125.33 79789 0.0217 Σu 10.0716 123.10 81234 0.0003 1 1 1 1 1 1 1 1 1 1 1 Only transitions up to approximately 10 eV are provided. Bold font indicates the first allowed transition at 170.3 nm from the HOMO to the LUMO+1 orbital and the mixture of three electronic transitions that make up the strongest band (see text for details). a to a degenerate LUMO pair of orbitals (MOs 17 and 18) corresponding to the A1Πg ← X1Σg+ state transition. This is what is predicted by a qualitative AH2 Walsh-type20 diagram for the simplest four-valence electron system. However, from the fvalues provided in Table 3, it is clear that while the lowest energy transitions occur at long wavelengths (in the region of the atomic zinc 4p 1P1 ← 4s1S0 transition) they have no oscillator strength due to being g−g forbidden. Consistent with experimental observations, the first transitions are predicted in the TD-DFT calculation to occur in the VUV. The calculated absorption of ZnH2 is shown by the stick spectrum in the lower panel of Figure S3. Shown also is the spectrum convoluted with a Gaussian function having a line width of 800 cm−1. As indicated in Table 3, the first allowed transition at 170.3 nm is from the HOMO to the LUMO+1 orbital. The strongest band, with a convoluted center at approximately 154 nm is, as specified in Table 3, a mixture of three electronic transitions. It consists of a degenerate 1Π state at 154 nm and a 1Σu state at 151.6 nm. The apparent disagreement between the intensities of the stick and convoluted spectra at 154 nm is an artifact of the presentation which arises due to the degeneracy of the excited 1Π states at this location. Shown in the bottom panel of Figure 6 are the absorption spectra recorded for ZnD2 both in a neat D2 solid (red curve) and in a D2-doped Ar matrix (gray curve). By comparing these curves with the convoluted DFT prediction21 (black curve, in the upper panel), it is evident that while the predicted bands are approximately in the same spectral region (the discrepancy is 4200 cm−1) the shapes are quite different. In particular a strong band consisting of two unresolved components of roughly equal intensity and a much weaker red band are predicted, while a broad structure consisting of a dominant band and a strong red shoulder are observed (red/gray traces in the lower Figure 6. Comparison of the TD-DFT predicted (convoluted) excitation energies with the absorption spectra of ZnD2. The dashed curves in the lower panel are those obtained by fitting four Gaussian curves to the recorded data. Their sum is shown by the gray curve overlaid on the experimental data, shown in red. The solid black traces in the upper panel are the stick and convoluted predicted spectra. To obtain a match with the recorded data, the gray traces were generated by blue shifting the original TD-DFT bands by 4200 cm−1 (shown by the solid black trace) and splitting the degenerate Π state by 2200 cm−1. The curve generated by summing the dashed curves is shown by the solid gray trace, and it can be compared with the experimental data in the lower panel. 9173 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178 The Journal of Physical Chemistry A Article doped-Ar system. However, as was observed in the ZnH2 system the predicted bands are all at lower energy than the recorded bands. In the comparison shown, the intensities of the TD-DFT calculated bands have not been altered. Other details of the fit are provided in the caption of Figure 7. panel). From the TD-DFT calculation it is known (see Table 3) that the transition at 154 nm is to a 1Π state, and it has been observed that the degeneracy of such states is often removed by the symmetry of the site accommodating the guest molecule in the solid matrix. In an attempt to account for the observed bands, the TDDFT transitions are all blue-shifted by 4200 cm−1, and the 1Π state has been split by 2200 cm−1, the result of which is shown by the four dashed curves in the upper panel of Figure 6. In so doing, a line width of 1050 cm−1 was used for each of the bands, and for all of them, a constant shift was applied. However, the predicted transitions intensities were not adjusted from the f-values obtained directly from the TD-DFT calculation and listed in Table 3. Addition of these four components (dashed curves) generates the solid black trace shown overlaid on the recorded ZnD2 spectra in the lower panel of Figure 6. The agreement is quite good; however, one aspect made evident in the comparison shown in Figure 6 is the existence of a lower energy band of ZnD2 at approximately 60 000 cm−1. This feature is not immediately obvious in the experimental spectra, in part due to the weakness of this band but also the poor signal-to-noise ratio. In Figure 7 the absorption spectra of CH3ZnH recorded in neat methane (red trace) and in CH4-doped Ar (gray trace) IV. DISCUSSION A. Metal Atom Spectroscopy. The absorption bandwidth (total fwhm 2100 cm−1) of the atomic 4p 1P1 ← 4s 1S0 transition in solid D2 is, as depicted in Figure 1, considerably greater than that of atomic zinc in the other light solids examined, namely, methane, and neon. Indeed it is much greater than what has been observed12 for the rare gases, Ar, Kr, and Xe. However, it is in line with values of 1500 and 1200 cm−1 measured for atomic magnesium22 and lithium,23 respectively, in solid hydrogen. On the other hand, the 3-fold splitting recorded in the Zn/D2 system is much better resolved than in the previous work done on Mg in H2, HD, or D2. From the spectra recorded for Li23 isolated in the solid hydrogens, it is not clear that 3-fold splitting exists in this system. The reason for the difference between zinc and the other metal atom systems is likely related to the smaller van der Waals radius of atomic zinc in the Zn·RG diatomics. Since the zinc atom−hydrogen molecule (Zn·D2) van der Waals distance is not yet known, one cannot definitively confirm this proposal. However, spectroscopic work on 1:1 metal atom/rare gas atom van der Waals complexes has revealed that, of the metals in question (Li, Mg, and Zn), the zinc-containing species have the shortest bond lengths. Given that the nearest neighbor (nn) distance24 in solid H2 is 3.79 Å, considerably greater than the nn distance in solid neon (3.09 Å) but very similar to that of solid Ar (3.76 Å), it is reasonable to propose that the same site occupancy will occur in solid H2 as solid argon. The results of a combined experimental/pairpotentials25 simulation26 suggest occupancy of atomic zinc in single vacancy (substitutional) site of argon. Thus, it is plausible that atomic zinc resides in a substitutional site in D2. This proposal is supported by the similarity of the 3-fold splitting revealed in the line shape fits presented in Figure S1 for Zn/Ar and Zn/D2 matrices. The much greater bandwidth of the latter system (742 cm−1) compared with Ar (412 cm−1, see Table 2) arises from the high phonon frequency of D2 which is a lowmass, quantum solid. As a result of the foregoing considerations of the size of atomic zinc, it is likely the site accommodating this guest is more symmetric than the distorted lattices accompanying isolation of the larger atoms Li and Mg if substitutional occupancies are also present for these guests. Moreover, the proposal of this site occupancy is consistent with the much more efficient isolation of atomic zinc in solid D2 than in solid neon as revealed in the absorption spectra shown in Figure 1. Multivacancy occupancy has been proposed11 for the site of isolation of atomic zinc in solid neon. B. Zinc Dideuteride Spectroscopy. The most likely candidate for the VUV absorption bands (shown in Figure 3) produced after atomic zinc photolysis in solid deuterium is zinc dideuteride, ZnD2. This molecule has been definitively identified in FTIR work5 of matrix-isolated atomic zinc in rare gases containing small amounts of hydrogen (and its isotopologues). The results obtained by Wang and Andrews4 for the reaction of atomic zinc in the solid para-hydrogen (and solid ortho-deuterium) confirm these assignments. The proposed mechanism for the formation of the ZnD2 molecule Figure 7. Similar comparison to Figure 6 but for CH3ZnH results. Three bands are sufficient to provide an acceptable fit of the recorded absorption bands. No splitting was used on the degenerate E state as it already has cylindrical symmetry in this C3v molecule. A constant shift does not allow the three convoluted bands to match the three recorded bands. In this system, a blue shift of 5200 cm−1 lines up the central band in the pure CH4 system. However, it is evident that this constant shift underestimates the location of the red band and overestimates the blue band. matrices are compared with a TD-DFT calculated electronic spectrum. When a comparison of both is made, it is evident that there is reasonable agreement between the broadened TD-DFT spectrum (black trace in upper panel) and the experimental spectrum. Thus three well-resolved peaks are predicted consisting of a strong blue band and two weaker red bands of equal intensity. As is evident in the lower panel of Figure 7, the recorded bands show this behavior as best illustrated in the 9174 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178 The Journal of Physical Chemistry A Article Table 4. Excitation Energies and Oscillator Strengths for Spin Allowed Transitions from the Ground X 1A1 State of CH3ZnHa state no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 a orbitals/coeff. 20 20 20 20 19 20 19 20 19 20 19 20 19 20 19 20 20 17 18 19 19 19 20 20 17 18 20 17 18 20 17 18 20 20 → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → → 21/0.69950 22/0.69950 23/0.68162 24/0.70006 22/−0.13028 26/0.69002 21/−0.13030 25/0.69001 22/0.65977 26/0.13528 21/0.65977 25/0.13530 23/0.48778 27/0.47850 23/0.47674 27/−0.43952 28/0.20483 22/0.30565 21/0.30487 23/−0.10968 24/0.31292 27/−0.12589 28/0.37893 29/0.17783 21/0.10735 22/0.108.19 31/0.68659 22/0.10735 21/−0.10840 30/0.68661 21/0.49798 22/−0.49783 28/−0.32111 29/0.61493 E (eV) λ(nm) E (cm−1) f E 1 E 1 A1 1 A1 1 A 5.6516 5.6516 6.2587 7.1196 7.6537 219.38 219.38 198.10 174.15 161.99 45583 45583 50479 57421 61732 0.0006 0.0006 0.1425 0.1784 0.0001 1 E 7.6537 161.99 61732 0.0001 1 E 7.9604 155.75 64205 0.2562 1 E 7.9604 155.75 64205 0.2562 1 A1 7.9907 155.16 64449 0.0603 1 A1 8.0550 153.92 64968 0.0846 1 A1 8.8590 139.95 71454 0.0002 1 E 8.8887 139.48 71695 0.0117 1 E 8.8887 139.478 71695 0.0117 1 A2 8.9005 139.30 71787 0.0000 1 A1 8.9127 139.11 71885 0.0467 state sym. 1 Only transitions up to approximately 9 eV are provided. is the insertion of the excited 1P1 state Zn atom into the D−D bond and the stabilization of the resulting intermediate in the low temperature solid. Gas-phase pump−probe studies9,27 of this reaction propose a similar species in the process of forming Zn−H + H. Moreover, the linear triatomics ZnH2 and ZnD2 have indeed been generated in the gas phase in a hightemperature furnace-discharge as observed1 in emission in the mid-IR region. The absorption bands of zinc dideuteride, ZnD2 (Figure 3) and methylzinc hydride, CH3ZnH (Figure 4), are featureless in the VUV spectral region. This characteristic arises because the excited states involved have all been shown to be dissociative. As indicated in Figures 6 and 7, the TD-DFT predicted spectra exhibit the right number of bands which are approximately in the correct spectral region for both species. However, they do require blue shifting of about 0.5 eV to achieve numerical agreement with the recorded data. The reason for this shift appears to be different for the two species. Thus, in CH3ZnH, a part of this effect arises from a matrix shift in the experimental spectra. This is revealed in the comparison of the pure methane data and the doped methane-Ar results, presented in the upper panel of Figure S2, which indicates the latter is blue-shifted against neat methane. The matrix shift arises due to the different interaction strengths between the guest species and the two host materials. As the site sizes and host polarizabilities of solid CH4 and Kr are very similar, the blue shift in Ar would indicate a repulsive interaction of the excited state guest molecule in its solid environment. This in turn would suggest CH3ZnH resides in a cramped site in argon compared with methane. In ZnD2 by contrast, the matrix shift appears to be nonexistent (see the lower panel of Figure S2). Even so, the theoretically predicted bands are red-shifted by about 4000 cm−1 (Figure 6). A possible reason for the difference is that the large Franck−Condon (FC) factors which accompany these dissociative electronic transitions are not included in the absorption band profiles generated in the present TD-DFT calculations. With the ultracold conditions used to record the absorption spectra, large FC factors will extend the TD-DFT calculated spectra to the blue-shift of the calculated vertical transition energies. The simple Gaussian convolutions done in the present simulations are evenly distributed around the electronic band origin and do not address this FC effect. Another possible reason for the 0.5 eV discrepancy could be due to inherent errors28,29 in the TD-DFT method itself. A rough indicator of the inaccuracy of a TD-DFT calculation is provided by the negative of the highest occupied Kohn−Sham (KS) orbital energy. Thus when the calculated excitation 9175 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178 The Journal of Physical Chemistry A Article Figure 8. A Walsh diagram calculated with the DFT method for HZnH. In performing this calculation, the Zn−H bond lengths were optimized for each specific angle. The four MO’s shown present the shapes of the HOMO and LUMO for the linear and bent (90°) geometries. Figure 9. A Walsh diagram calculated for CH3ZnH with the DFT method as a function of the C−Zn−H angle. In doing this calculation, all of the bond lengths were optimized for each specific angle. 9176 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178 The Journal of Physical Chemistry A Article shape of the energy curve of the stabilized LUMO (MO17) of ZnH2 (CH3ZnH, MO21) becomes almost linear after only a small (approximately 15°) angle deviation from 180° while the shape of the destabilized HOMO energy, MO16 (MO20), is curved over the entire angle range. As Figures 8 and 9 respectively reveal, ZnH2 and CH3ZnH both show this same behavior. The curved behavior of the HOMO energy can be traced back to the angle dependence of the overlap between the Zn atom p-orbital and the H-atom s-orbital. As such, it will simply exhibit a cosine dependence as is well-illustrated in Figure S4. By contrast the cosine dependence of the LUMO is limited to small angle deviation from the linear geometry. Thereafter, the energy of this orbital is stabilized by the formation of the new H−H bond. This contributing effect is usually not considered in textbook descriptions of AH2 Walsh diagrams. energies approach this threshold, the predicted values are likely to be underestimated relative to experimental results. Following this rough accuracy guide, the negative of the HOMO energy of ZnH2 (CH3ZnH) is 63445.7 (57855.7) cm−1, and the first allowed transition is, as listed in Tables 3 and 4, less than this, having a value of 58730 (50479) cm −1 . From these comparisons we expect that, while not an ideal situation, the B3LYP functional (with the large basis set used) should allow the TD-DFT method to provide a reasonable prediction of the excitation energies. Indeed, the underestimate of the predicted transition energies is less than 6% (0.5 eV) on the allowed transitions in the 8 eV region. It should be mentioned that the TD-DFT method also underestimates the atomic zinc 4p 1P1 ← 4s 1S0 transition but by a somewhat smaller amount. Thus the gas phase value is known17 to be 46745.413 cm−1, while the predicted value is 45531 cm−1corresponding to an underestimate of 2.6%. C. Photochemistry. The dissociative activity exhibited by the VUV bands of ZnD2 and CH3ZnH upon photolysis is consistent with the featureless nature of these absorption bands. This behavior can be related to the bonding described for a four-electron AH2 system by a Walsh diagram. Thus the ground state is, as observed experimentally, predicted to be linear in this simple bonding model. The first excited states on the other hand are expected to be bent due to strong stabilization of the LUMO orbital upon bending. To investigate the observed photodissociation of ZnH2 and CH3ZnH, we explored the evolution of the orbital energies of both molecules from the linear to bent geometries. In analyzing this we present in Figures 8 and 9 graphs of the ZnH2 and CH3ZnH MO energies between 180° and 90°, calculated in 5° increments. Also included are the MO shapes as determined in Gaussian 03. From Figure 8, it can be seen that when ZnH2 is in the ground state and the HOMO is occupied, the linear form is more stable than the bent. When it absorbs at 154 nm, one of the valence electrons in MO15 (HOMO-1) is promoted to the LUMO (MO’s 17/18) as indicated in Table 3. In the excited state it can be seen that the bent form of ZnH2 is actually more stable than the linear. The angle dependence of the energy of MO17 is the reason for the change in geometry. MO17 of ZnH2 in the linear geometry consists of a nonbonding p-orbital centered on the Zn atom, as depicted on the top left of Figure 8. The bent form contains the same nonbonding p-orbital, but in this geometry (see top right of Figure 8) the s-orbitals overlap between the two hydrogen atoms. Arising from this change is the formation of an H−H bond. Moreover, as there is also no electron density over the Zn and H atoms, breakage of the Zn−H bond results. It is this evolution of MO17 which leads to the photodissociation of ZnH2 and the formation of atomic zinc and molecular hydrogen. A similar but more complex behavior was found in methylzinc hydride as shown in Figure 9. However, just as in the ZnH2 system, at 90° MO21 of CH3ZnH shows the reformation of a C−H bond from the hydrogen that was bonded to the zinc atom. A node is also evident (see top right of Figure 9) between the zinc atom and the CH3···H moiety. The angle dependence found in the TD-DFT calculations of the LUMO energies is identified as responsible for the clean, concerted dissociation of the VUV bands in the case of both molecules. This is the classic behavior predicted by a simple four-valence electron AH2 Walsh-type diagram. In contrast to the aforementioned behavior, correctly anticipated by a Walsh diagram, it is noteworthy that the V. CONCLUSIONS The absorption spectra of ZnD2 and CH3ZnH have been recorded for the first time in the vacuum-UV range. These “inserted” species were produced from the efficient reaction of excited 1P1 state atomic zinc with the molecular precursors D2 and CH4. Emission of the 1P1 state was completely quenched in both solid D2 and CH4; however, 3P state luminescence of atomic zinc was observed in the Zn/CH4 system. The latter observation points to the involvement of the spin triplet state in the relaxation of the lower symmetry CH3ZnH system as it evolves into the C3v ground state geometry. This state appears to play no role in the stabilization of ZnD2. TD-DFT calculations provide reliable descriptions of the excited states responsible for the VUV absorptions of ZnH2 and CH3ZnH; however, all of the predicted transition energies are underestimates of the observed bands. Large Franck−Condon factors may partly explain the blue-location of the observed bands relative to the vertical excitations generated in the TD-DFT calculations. The observed photodissociation of these molecules has been modeled successfully with Walsh-type diagrams generated from the DFT calculations. ■ ASSOCIATED CONTENT S Supporting Information * Lineshape analyses of absorption profiles for Zn/D2, Zn/CH4, and Zn/Ar (Figure S1), a comparison of the VUV absorption bands produced in Zn/D2 and Zn/CH4 systems (Figure S2), a comparison of the TD-DFT predictions for the electronic transitions of CH3ZnH and H−Zn−H (Figure S3), DFT calculated overlap dependence of the Zn 4p orbital and H atom 1s orbital (Figure S4), comparison of band origins (Table S1), and band positions extracted by free-fitting Gaussan curves to the recorded absorption data (Table S2). This material is available free of charge via the Internet at http://pubs.acs.org. ■ AUTHOR INFORMATION Corresponding Author *E-mail: john.mccaff[email protected]. Notes The authors declare no competing financial interest. ■ ACKNOWLEDGMENTS We would like to acknowledge Dr. Peter Gürtler and Dr. Sven Petersen for technical assistance with the experimental work done at the HASYLAB synchrotron. This research was funded 9177 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178 The Journal of Physical Chemistry A Article in part by the European Union, TMR, “Access to Large Scale Facilities” Programme. C.H. also gratefully acknowledges receipt of a Hume Ph.D. studentship, at NUI-Maynooth. ■ (19) Bracken, V. A.; Guertler, P.; McCaffrey, J. G. Spectroscopy and Photodissociation of Dimethylzinc in Solid Argon. 1. Vacuum UV Luminescence Detection/Synchrotron Radiation Photolysis. J. Phys. Chem. A 1997, 101 (51), 9854−9862. (20) Gimarc, B. M. Molecular Structure and Bonding; Academic Press: New York, 1979 Chapter 3. (21) In Figure 9 theoretical predictions for ZnH2 are used in a comparison with the experimental data recorded for ZnD2. This is possible because the electronic spectroscopy is not dependent on the masses of the isotopes. (22) Presser, N. Spectroscopy of magnesium in hydrogenic matrixes. Chem. Phys. Lett. 1992, 199 (1−2), 10−16. (23) Fajardo, M. E. Matrix-isolation spectroscopy of metal atoms generated by laser ablation. II. The lithium/neon, lithium/deuterium, and lithium/hydrogen systems. J. Chem. Phys. 1993, 98 (1), 110−18. (24) Silvera, I. F. The solid molecular hydrogens in the condensed phase: fundamentals and static properties. Rev. Mod. Phys. 1980, 52 (2, Pt. 1), 393−452. (25) Kerins, P. N.; McCaffrey, J. G. A pair potentials study of matrixisolated atomic zinc. I. Excited 1P1 state dynamics in solid Ar. J. Chem. Phys. 1998, 109 (8), 3131−3136. (26) Healy, B.; McCaffrey, J. G. Simulation of atomic cadmium spectroscopy in rare gas solids using pair potentials. J. Phys. Chem. A 2000, 104 (16), 3553−3562. (27) Umemoto, H.; Tsunashima, S.; Ikeda, H.; Takano, K.; Kuwahara, K.; Sato, K.; Yokoyama, K.; Misaizu, F.; Fuke, K. Nascent internal state distributions of ZnH(X2Σ+) produced in the reactions of Zn(4 1P1) with some alkane hydrocarbons. J. Chem. Phys. 1994, 101 (6), 4803−8. (28) Casida, M. E.; Jamorski, C.; Casida, K. C.; Salahub, D. R. Molecular excitation energies to high-lying bound states from timedependent density-functional response theory: characterization and correction of the time-dependent local density approximation ionization threshold. J. Chem. Phys. 1998, 108 (11), 4439−4449. (29) Because the accuracy of the method depends on the exchangecorrelation functional used, TD-DFT may not provide exact transition energies. As is now well-known, the exchange-correlation functionals work very well for ground states and for low-lying valence excited states. However, for high-lying diffuse excited states TD-DFT employing conventional functionals may be less reliable, and the excitation energies can be substantially underestimated. This is because the exchange-correlation potentials generated from these functionals decay rapidly (exponentially) rather than the gradual (1/r) asymptotic decay of the true potential. REFERENCES (1) Shayesteh, A.; Gordon, I. E.; Appadoo, D. R. T.; Bernath, P. F. Infrared emission spectra and equilibrium bond lengths of gaseous ZnH2 and ZnD2. Phys. Chem. Chem. Phys. 2005, 7 (17), 3132−3142. (2) Sebald, P.; Vennekate, H.; Oswald, R.; Botschwina, P.; Stoll, H. A theoretical study of ZnH2: a case of very strong Darling-Dennison resonance. Mol. Phys. 2010, 108 (3−4), 487−499. (3) Flory, M. A.; Apponi, A. J.; Zack, L. N.; Ziurys, L. M. Activation of Methane by Zinc: Gas-Phase Synthesis, Structure, and Bonding of HZnCH3. J. Am. Chem. Soc. 2010, 132 (48), 17186−17192. (4) Wang, X.; Andrews, L. Infrared Spectra of Zn and Cd Hydride Molecules and Solids. J. Phys. Chem. A 2004, 108 (50), 11006−11013. (5) Greene, T. M.; Andrews, L.; Downs, A. J. The Reaction of Zinc, Cadmium, and Mercury Atoms with Methane: Infrared Spectra of the Matrix-Isolated Methylmetal Hydrides. J. Am. Chem. Soc. 1995, 117 (31), 8180−7. (6) Bracken, V. A.; Legay-Sommaire, N.; McCaffrey, J. G. Spectroscopy and Photodissociation of Dimethylzinc in Solid Argon. 2. FTIR Detection/ArF Laser Photolysis. J. Phys. Chem. A 1997, 101 (51), 9863−9869. (7) Breckenridge, W. H. Activation of H-H, Si-H, and C-H Bonds by nsnp Excited States of Metal Atoms. J. Phys. Chem. 1996, 100 (36), 14840−14855. (8) Breckenridge, W. H.; Wang, J. H. Dynamics of the reactions of zinc(4s4p3P1) with hydrogen, hydrogen deuteride, and deuterium: rotational and vibrational quantum-state distributions of zinc hydride (zinc deuteride) (ZnH (ZnD)) products. J. Chem. Phys. 1987, 87 (5), 2630−7. (9) Umemoto, H.; Tsunashima, M.; Tsunashima, S.; Ikeda, H. Nascent vibrational state distributions of ZnH(X2Σ+) and ZnD(X2Σ+) produced in the reactions of Zn(41P1) with H2 and D2. Laser Chem. 1995, 15 (2−4), 123−9. (10) Martinez-Magadan, J. M.; Ramirez-Solis, A.; Novaro, O. MCSCF + MR-CI study of the interaction of atomic zinc, and zinc(1+) and zinc(2+) ions with the hydrogen molecule. Chem. Phys. Lett. 1991, 186 (1), 107−12. (11) Healy, B.; Kerins, P.; McCaffrey, J. G. Metal atom (Zn, Cd and Mg) luminescence in solid neon. Low Temp. Phys. 2012, 38 (8), 679− 687. (12) Bracken, V. A.; Gurtler, P.; McCaffrey, J. G. Luminescence spectroscopy of atomic zinc in rare-gas solids. I. J. Chem. Phys. 1997, 107 (14), 5290−5299. (13) Collier, M. A.; McCaffrey, J. G. Luminescence spectroscopy of matrix-isolated z6P state atomic manganese. J. Chem. Phys. 2005, 122 (18), 184507/1−184507/15. (14) Frisch, M. J., et al. Gaussian-03, Revision E.01; Gaussian, Inc.: Wallingford, CT, 2004. (15) Murray, C.; Dozova, N.; McCaffrey, J. G.; Shafizadeh, N.; Chin, W.; Broquier, M.; Crepin, C. Visible luminescence spectroscopy of free-base and zinc phthalocyanines isolated in cryogenic matrixes. Phys. Chem. Chem. Phys. 2011, 13 (39), 17543−17554. (16) Alikhani, M. E. Theoretical study of the insertion reaction of zinc, cadmium, and mercury atoms with methane and silane. Chem. Phys. Lett. 1999, 313 (3), 608−616. (17) NIST Atomic Spectra Database, Version 4; National Institute of Standards and Technology: Gaithersburg, MD, 2013; http://www.nist. gov/pml/data/asd.cfm (available: May 2013). (18) Because of the high repetition rate of the SR excitation pulses, direct measurement of the Zn/CH4 near-UV emission decaytimes was not possible. Even for “single-bunch” mode operation, when the repetition rate of the storage ring is reduced to 1.042 MHz, this was still too high indicating a very long-lived emission. Accordingly it is inferred that the decay time of the near-UV emissions must be longer than 10 μs. 9178 dx.doi.org/10.1021/jp404695a | J. Phys. Chem. A 2013, 117, 9168−9178